Phylogenetic Relationships among European Polistes and the Evolution ofSocial Parasitism (Hymenoptera: Vespidae, Polistinae) James M. CARPENTER Department ofEntomology, American Museum ofNatural History, Central Park West at 79th Street, New York, NY 10024, U.S.A. ABSTRACT Cladistic analysis of the European species ofPolistes is used to investigate the evolution .of social parasitism in the genus. The three species of social parasites (formerly the genus Sulcopolistes) are all inquilines: lacking a worker caste, and dependent on uSUIping the colony ofa host species to obtain a worker force. El:vlERY'S Rule states that social parasites are more closely related to their hosts than to any other species. Previously published allozyme (CARPENTER et aI., 1993) and mtDNA (CHOUDHARY et aI., 1994) data did not support this hypothesis, but did not resolve relationships among the species of social parasites. Morphological characters are adduced which resolve the phylogenetic relationships among these three species, and the combination of the morphological and molecular data sets largely resolves the relationships among all nine European species. Cladistic optimization oftraits associated with social parasitism on the resulting cladogram shows: (1) El:vlERy's'Rule is rejected; (2) the scenario proposed by TAYLOR (1939) for the evolution of social parasitism is not supported either. The predatory behavior ofthe inquiline P. atn"mandibularis, with separate "supply" and "nursery" nests, is evidently secondary, as .. is its initially passive invasion behavior. - La phylogenie des Polistes d'Europe et revolution du parasitisme social (Hymenoptera: Vespidae, Polistinae) L'analyse cladistique des especes de Polistes europeennes est utilisee pour etudier l'evolution du parasitisme social dans Ie geme. Les trois especes de parasites sociaux (autrefois Ie genre Sulcopolistes) sont toutes « inquilines» : elles n'ont pas de caste ouvriere et s'approprient celle de la colonie d'une espece hOte. Le regIe d'El:vlERY afIirme que les especes de parasites sociaux sont plus proches parents de leurs hOtes que de toute autre espece. Des analyses basees sur les allozymes (CARPENTER et aI., 1993) et l'ADN mitochondrial (CHOUDHARY et al., 1994) n'ont pas confirme cette hypothese, mais n'ont pas resolu non plus les relations entre les especes parasites. L'addition de caracteres morphologiques et la combinaison des donnees morphologiques et moleculaires ont permis de resoudre respectivement 1es relations phylogenetiques entre ces trois especes, et entre les neufespeces europeennes. L'optimisation sur Ie cladogranune de traits associes au parasitisme social montre que : (1) Ie regIe d'El:vlERY est refutee ; (2) Ie scenario d'evolution du parasitisme social propose par TAYLOR (1939) est egalement refute. Le comportement predateur de l'inquiline P. atn"mandibulan's, qui maintient des nids separes pour les provisions et pour Ie couvain, est al'evidence secondaire, de meme que son comportement d'invasion initiale passive. CARPENTER, J. M., 1997. - Phylogenetic relationships among europeanPolistes and the evolution ofsocial parasitism (Hymenoptera: Vespidae; Polistinae). In: GRANDCOLAS, P. (ed.), The Origin of Biodiversity in fusects: Phylogenetic Tests of Evolutionary Scenarios. Mem. Mus. natn. Hist. nat., 173: 135-161. Paris ISBN: 2-85653-508-9. 136 1. M. CARPENTER: EVOLUTION OFSOCIAL PARASITISMINPOLISTES INTRODUCTION The social parasites known as inquilines are among the most intngumg outcomes of evolution in paper wasps. Inquilines have no worker caste and cannot build nests; they must invade the colony-of-iIiostspeCiesandsuppliiiirrhe-queen to obtain workersto-rea.r1lieir brood.·--­ Striking morphological differences occur between host and parasite (ef Figs 1-2). This form of behavior is rare in paper wasps, having been found in just three species in one genus, Polistes. Polistes is a cosmopolitan genus, with more than 200 described species (CARPENTER, 1996b); the three social parasites (atrimandibularis, semenawi and suleifer) are found only in Europe and the Mediterranean Region. The inquiline behavior has long been known (WEYRAUCH, 1937), but recent years have seen a profusion of studies, particularly by workers at the University of Florence (for a review see CERVO & DANI, 1996). These studies have revealed remarkable phenomena, such as simultaneous domination by atrimandibularis ofseveral colonies ofthe host biglumis, only one ofwhich serves for parasite brood rearing while the larvae ofthe other nests serve as a source of food (CERVO et aI., 1990c), or change in the composition of the chemical signature ofatrimandibularis to match that ofbiglumis at the time ofemergence ofhost workers (BAGNERES et al., 1996). This wealth of new information has stimulated intense interest in evolutionary explanations ofaspects ofsocial parasitism. In this paper I apply cladistic analysis to investigate the evolution of social parasitism in Polistes. I first review evolutionary explanations that have been advanced for the social parasites, and how cladistic tests may be constructed for such hypotheses. I then present the first cladistic analysis of interrelationships among the inquilines and related species to be based on morphology. These characters are also combined with previously published molecular data, and analyzed simultaneously. Behavioral features associated with social parasitism are optimized on the resulting cladogram, to test various evolutionary hypotheses. EMERY's Rule The classical explanation that applies to the origin of socially parasitic species is the hypothesis known as EMERY'S Rule (after EMERY, 1909). EMERY's Rule is that social parasites are more closely related to their host species than to any other species. This is usually interpreted to mean that social parasites evolved directly from their hosts, either by sympatric speciation (e.g. WEST-EBERHARD, 1986; BUSCHJNGER, 1986, 1990; BOURKE & FRANKs, 1991) or not (e.g. WILSON, 1971). Alternatively, social parasites may exploit similar recognition systems in closely related species (CARLIN, 1988). However, previous cladistic analyses do not support EMERY's Rule in Polistes. CARPENTER et al. (1993) presented an allozyme data set for the three species of social parasites, their four host species (see Table 1) and two outgroup species. The 27 loci were subjected to three different coding procedures, which led to different results for each method, but under none of the codings were the social parasites most closely related to their hosts. CHOUDHARY et al. (1994) analyzed a 386 base pair sequence from the mitochondrial16S rRNA gene. Their cladistic results were less ambiguous, with two cladograms, differing only in the interrelationships among the social parasites. The social parasites were a monophyletic group. Thus, phylogenetic analysis of two independent sources of evidence, by rejecting close relationship ofparasite to host, rejected EMERY's Rule. PHYLOGENETIC TESTS OF EVOLlTTIONARY SCENARlOS 137 FIGS 1-2. - Frontal view. 1: Polistes dominulus; 2: Polistes atrimandibularis. The magnification is 21x. TABLE 1. - Host-parasite relations among European Polistes species. CERVO & DAN! (1996) report experimental introduction ofatrimandibularis into the nests ofnimphus, but this is unknown to occur naturally. Parasite Host atrimandibularis biglumis, ga//icus semenowi dominulus, nimphus sulcifer dominulus TAYLOR'S Scenario EMERY's Rule also played a role in development of an evolutionary scenario for the origin and elaboration of social parasitism in wasps, that of TAYLOR (1939). TAYLOR observed a case ofnest usurpation ofthe vespine species Vespula vidua by V. squamosa, then considered closely related. He then suggested a sequence of behavioral changes from free living to inquiline. The scenario consisted of four stages: (1) Intraspecific, facultative, temporary parasitism, (2) Interspecific, facultative, temporary parasitism, (3) Interspecific, obligate, temporary parasitism, (4) Interspecific, obligate, permanent parasitism. Intraspecific nest usurpation thus evolved into inquiline behavior, with the loss of a worker force produced by the usurping queen. The existence of intraspecific nest usurpation was well known in vespines (JANET, 1903), as were inquilines (e.g. WEYRAUCH, 1937), and TAYLOR's scenario became generally accepted (e.g. WILSON, 1971; MATIHEws, 1982; GREENE, 1991). The applicability ofthis hypothesis to polistine social parasites has been questioned by CARPENTER et 138 1. M. CARPENTER: EVOLUTION OF SOCIAL PARASITISM INPOLISTES al. (1993) and CHOUDHARY et ai. (1994), who observed that whereas the first stage in TAYLOR's scenario was common in paper wasps, the second stage was rarely reported, and the third stage is unknown (see review in CERVO & DANI, 1996). Nevertheless, the general notion of evolution ofusurpation into inquilinism was accepted. Recent hypotheses The recent increase in knowledge of the behavioral ecology of the inquilines has been accompanied by novel hypotheses on the evolution of features of their biology, some newly reported. Notable examples discussed by CERVO & DANI (1996) include observations on the distribution of parasites and length of the colony cycle in one host, timing of usurpation by the parasite, the tactics employed by the parasite in dominating the host queen, and the predatory behavior ofatrimandibularis. These ideas are briefly summarized here. The Florence group has shown that inquilines are moderately abundant at the foot of mountain ranges. The hosts